I invite you to consider a rather dry problem. I ventured to select this topic because it has lately been my duty to occupy myself with certain legal writings, which, perhaps, took me a little beyond my depth. They touched, however, problems which are common to the lawyer and to the moralist. Although not a lawyer, I am interested in some moral problems which have also a legal aspect: What I propose to do this evening is, to consider certain questions which lie in the region common to both provinces of inquiry, and especially this question: What is the true ethical theory of punishments inflicted by the criminal law? How, and in what sense, are they to be regarded as just? There is, obviously, a relation between the two codes—moral and legal. Murder is both a sin and a crime: a breach of the moral law, and of the laws of every civilised country. Yet, there is one broad and deep distinction between the two systems of law. The moral law is essentially concerned with a man's motives. To say that a man's conduct is wicked, is necessarily also to say that it is the action of a bad man, or due to evil passions. Murder is wicked, as it is the manifestation of the murderer's hatred of his neighbour. The criminal law, on the other hand, has to deal, in the first instance, with the external facts. It contemplates, primarily, what a man does, not what he is. It does not attempt to punish every man who hates his neighbour, but every man who has, in fact, killed, whether the action springs from hatred or some other motive. Every one who deliberately kills, unless the act falls under certain definite exceptions, is guilty of murder. This, of course, does not imply that the moral aspect is of no account. The exceptions are so arranged that the legal classification corresponds roughly to the moral classification. Under certain exceptions, killing is regarded as justifiable homicide, and under others, it is only manslaughter, and, therefore, receives none, or a slighter penalty. The coincidence between the codes may thus be very close. In ninety-nine cases out of a hundred the action condemned by the criminal law will be condemned by the moralist. The man who is legally guilty of murder is also, almost invariably, guilty of a great moral offence. Although, again, the moral law applies to large classes of conduct, which are not within the cognisance of the criminal law, it is, at least, plainly desirable that the criminal law should condemn nothing which is not also morally wrong. The sway of the moral law is universal; it applies to all conduct, and, of course, to the conduct of legislators and judges: they and the law which they define and apply should be consistent with the general law of right and wrong. They and all of us are bound not to make virtue more difficult nor vice easier.

But, further, the questions as to the relations between the two codes arise in various directions. It is obvious that the criminal law has to employ very rough and ready methods. It cannot estimate, with any accuracy, the degree of immorality implied by any given action. It cannot, and it does not attempt to, look closely into the secrets of a man's heart. It cannot inquire, as a rule, how far a man's crime is the result of bad education or bad surroundings; how far it implies thorough corruption or only superficial faults of temper, or a misunderstanding of some fact or doctrine. It cannot take into account a number of metaphysical or psychological considerations which are connected with the theory of moral responsibility. To settle such points you would have to empanel a jury of philosophers, and the only thing of which you could be certain would be, that such a jury would never agree upon a verdict. Again, there are whole classes of virtues and vices with which the criminal law is not concerned. Ingratitude, to take the common example, is a grave vice, but one which it would be absurd to punish legally. Not only would such an attempt involve impossible inquiries, but the attempt would be self-defeating. If the duty of gratitude to a benefactor were turned into a legal obligation, gratitude proper would cease to exist. To confer a benefit would be the same thing as to acquire a right to repayment. A man who allows his best friend to starve, or to go to the workhouse, may be, morally, far worse than a thief; but you could not punish him legally, without adopting a principle which, even if practicable, would, so far as it operated, be destructive of all disinterested friendship. The law, again, can deal only with criminals who are found out. What proportion they may bear to the whole class of moral offenders is not discoverable; but it is, at least, safe to say that, for every man whom you convict of a crime, you must leave unpunished, because undetected, another sinner who is equally deserving of punishment. And, finally, it is apparently impossible to say, upon any intelligible grounds, what should be the proportion between crime and punishment. How many years' imprisonment does a man deserve for putting out his neighbour's eye? I do not see how such a rule of three can be stated. The good old theory of an eye for an eye and a tooth for a tooth, seems to suggest a possible criterion. But it was difficult to carry out. Deloraine, in the Lay of the Last Minstrel, has, as he points out, killed Musgrove's brother; but, on the other hand, Musgrove has killed Deloraine's nephew, and, besides, got a thousand marks ransom out of Deloraine himself. Is the account to be regarded as accurately balanced? Is one brother just equal to a nephew plus a thousand marks? The theory, of course, is an application of an inappropriate analogy. If we regard crime simply as a case of private injury, we may say that it is fair that the wrong-doer should restore the thing that he has taken, and so put matters where they were before. But this is obviously to take a view which is quite inapplicable in most cases, and in all cases becomes inadequate when we take the moral view, and regard crime as an offence against society—not simply as a wrong to another individual.

For such reasons, it is apparently impossible to say that a legal punishment can be just, in the full sense in which the moralist would use the words. No doubt we may say,—and we wish that we could always say,—that a man "deserves" what he has got; and that implies that we recognise as desirable some satisfaction to our sense of justice. And, of course, too, we demand that justice should be done in another sense of the word; that the case, for example, should be impartially investigated; that a man should not be punished severely because he is poor, or because he is unpopular, or let off easily because he is a private friend of the judge. Such demands mean that justice should not be perverted by applying irrelevant considerations; but they leave our previous questions untouched. The criminal law, from its nature, cannot impose equal penalties upon all men who are equally wicked; but only upon those who have made themselves liable: and that always involves elements of accident; it cannot take into account at all some of the elements upon which the depth of moral depravity essentially depends; and it is, at least, very difficult to say what specific meaning can be given to the proportion between crime and the suffering imposed upon the criminal.

If, then, the legislative action must, of necessity, be very imperfect from the moral point of view, we may try what will be the effect of dismissing the moral question altogether, or, at least, reducing it to a secondary place. We may, that is, consider crime not in so far as immoral, but in so far as mischievous. Here we have the doctrine worked out very consistently by Bentham and his followers. Pain, they said, is an evil, the only evil; pleasure, a good, and the only good. To inflict needless pain—pain which does not cause a balance of pleasure—upon any one, be he a good man or be he a bad man, is, so far, wrong. For the same reason, it is justifiable, and, indeed, right, to inflict pain, so far as it prevents some greater evil. Hence, you should punish criminals just so far as the pain which you inflict is less than the pain which you prevent. It is wrong to give a single useless pang even to the worst of men. If (according to a sentiment attributed to Bentham) a fine of five shillings would prevent a man from committing murder, it would be wrong to fine him seven shillings and sixpence. This gives a justification of punishment, in so far as deterrent. It is obviously connected with another doctrine. A man is the best judge of his own pleasures and pains. Therefore, in so far as a man's actions affect himself alone, they are not to be forbidden by the law. We may think them bad or degrading; but so long as they do not affect others, the fact that a man chooses them is a proof that they give him pleasure; and we shall, therefore, only diminish the sum of happiness by interfering. Now, it is plain that this distinction does not draw the line between what is morally bad or good. Every habit which affects a man's own character, affects, also, his capacity to fulfil his duties to others. But this theory overlooks immorality, except so far as it happens to involve certain extraneous consequences. We are, upon this showing, to punish a criminal precisely in the same spirit as we are to abate a nuisance. The thief is to be suppressed, as we are to extirpate a mischievous weed, and to be suppressed by just as much severity as is required for the purpose. The drunkard, so long as he confines himself to making a beast of himself in his own room, does his neighbours no direct injury, and must be left to enjoy the pleasure which is shown, because he chooses it, to be a pleasure to him. Of this theory, it may, I think, be said that, however imperfect, it is tolerably consistent, and, moreover, that it undoubtedly does express one legitimate end of punishment. There can be no doubt, that is, that the punishment of murderers may be rightly defended, among other grounds, at any rate, on the ground that it discourages the practice; though we may not fully agree with the famous saying of the judge, "You are not hanged for stealing sheep, but hanged in order that sheep may not be stolen". And, further, though there are various difficulties about the distinction between "self-regarding" and "extra-regarding" conduct, we must also, I think, allow, in general terms, that the fact that a man's conduct has a direct and assignable influence upon his neighbour's happiness, must always be one reason, and, frequently, the only sufficient reason, for suppressing it by legal penalties.

This doctrine of simple deterrence, however, seems, to most critics, to be insufficient. It omits the moral element too completely. When a man is punished for some revolting offence, we are not simply providing him and his like with reasons for abstaining in future. We are, as a fact, exposing him to infamy, sometimes more painful to bear than the immediate penalty, and are thus, in fact, invoking the sanction of the moral sentiment. Therefore, it is urged, we must still, whether we like it or not, be moralists. The purely utilitarian argument has omitted one element of the calculation. The punishment not only deters offenders, but gratifies the feeling of resentment to moral indignation, which has been approved by many moralists. Hence, it is urged, besides the deterrent theory, we must make room for the vindictive theory. It is legitimate and right to hate crime, and, therefore, to hate criminals; and legal punishments are defensible, not merely as adding to the motives for refraining from crime, but as gratifying the desire for revenge, which, in early ages, was assumed in the rude modes of putting down violence, and which, even now, should be not eradicated but confined within legal channels and directed towards the desirable ends.

Postponing, for the present, a consideration of this proposed emendation, let us consider, a little more closely, the objection made to the theory of deterrence. In what way does it come into direct conflict with a moral theory of punishment? It looks upon immorality as mischievous, or as diminishing happiness; and upon the utilitarian view immorality means the diminution of happiness. Now, without discussing ultimate moral questions, I may assume that, for practical purposes, this seems to be a sufficiently tenable position. After all, we admit, to whatever school we belong, that crime is mischievous, and, whatever deeper meaning may be assigned to it, may be considered in that light by the legislator. He cannot—certainly he ought not to—forbid actions which do no harm to anybody, or which nobody, at the time and place, feels to be injurious to happiness. Even, therefore, if utilitarianism be unsatisfactory as an ultimate theory, it may represent adequately the point of view of the practical legislator. He tries to suppress violence and fraud because, as a fact, they cause what their victims unanimously agree to be painful consequences; and he need not look any further for a reason. People, it is said, have very different standards of pleasure. Still, we all dislike having our throats cut or our pockets picked; and that fact supplies a sufficient ground upon which to base the whole criminal law. When we go a little further, a point of divergence may be noticed, a short consideration of which may help to clear the case. Let us assume the legitimate end of all punishment to be deterrence. It will follow, that we must annex as a consequence to crimes an adequate counterpoise, and a counterpoise not more than adequate to the criminal's motives. The fine to be paid must be just sufficient to prevent the transgression. Now, it has been urged, this necessarily implies a conflict with morality. The degree of moral guilt implied in a given crime varies inversely as the temptation. The greater the inducement to the offence, the less the wickedness shown in committing the offence. A man may have enough virtue to refrain from a gratuitous injustice, although he has not virtue enough to resist a large bribe, or the threats of a man in power. But, if the legislator is to provide simply a counterpoise, he will have to follow the opposite rule. The greater the temptation, the greater must be the force of the motive which must be added to counterbalance the temptation. If there be a crime by which a man might make a million of money, you must, if you would prevent it, hold out the prospect of such pains as would, in his estimation, be cheaply avoided at the sacrifice of a million; or, making allowance for the uncertainty of detection, by the sacrifice of more than a million. But if, by the same crime, he only got a five-pound note, the prospect of paying a hundred pounds in case of detection might be a sufficient preservative of his honesty. Yet, the man who is tempted by the million gives less proof of dishonesty than the man who commits the same crime for a paltry five pounds. Therefore the punishment must be increased, as the wickedness is less.

I must first set aside one ambiguity which perplexes this argument. When we speak of a temptation as varying, we may mean one of two very different things. To say that I am more "tempted" than you to commit a given crime, may mean that the gain expected by me is itself greater; or, it may mean that I am more predisposed to the crime. I may be more tempted, let us say, to poison my uncle than you are to poison yours. That may mean that my uncle is a rich old sinner and I am his heir, whereas your uncle is a poor saint and you will get nothing by his death. Or it may mean that I am more tempted because, our uncles being alike, I am spiteful, and you affectionate, by nature. In the first case, to say that I am under the stronger temptation would, perhaps, tend to alleviate the gravity of my crime; in the second, it would simply be another way of saying that I was the greater brute. In both cases, of course, it is true that the greater temptation would require the greater counterpoise. In one case, this only means that the worse the man, the stronger the restraints which he requires; and, if you could make different laws for bad men and good, it would follow that the bad would require the heaviest penalties. But this does not conflict with the moral view. It is no excuse for a murderer to say, "I am so bloodthirsty that I really could not help murdering". No contradiction to morality arises from punishing his crime more severely. In the other case alone,—the case in which we made distinctions founded upon the difference of surrounding circumstances,—it is true that we should, from the point of view of simple deterrence, require heavier penalties where the temptations were greater, and, therefore, the intrinsic malevolence proved to exist less.

For most purposes, this argument seems to have very little practical application. The law is made for people in general; we cannot have one law for bad men and another for good; partly because good and bad people do not carry about tangible marks of their quality written upon their faces. No doubt, indeed, the atrocity of a crime is recognised, if not by the general law, by the nature of the sentence. An assault may show unnatural ferocity or merely a rather excessive warmth of temper; and, though the offence may be forbidden under the same clause of the criminal law, the judge may be empowered to give sentences of varying severity, varying more or less according to the moral depravity implied. So far, the worst offences (in a moral sense) get the heaviest punishment; and the deterring influence is rightly exerted by proportioning the penalty to the temptation, that is, to the predisposition to crime. The other case, again, requires some qualification. It is not true, as an absolute proposition, that the criminality is always, or generally, diminished, in proportion to the greatness of the temptation; for we must remember that both the temptation and the crime will generally be greater in proportion to the amount of mischief inflicted. It is more tempting, no doubt, to appropriate a thousand pounds than a shilling; but we cannot infer that the man who takes the larger sum is, therefore, less wicked; that he has a conscience which would have kept him honest under the smaller temptation, and has only yielded to the greater. Compare, for example, the case of the petty pilferer who appropriates my watch, with the case of the man of business who appropriates securities worth many thousand pounds and ruins widows and orphans by the dozen. We should all agree, I imagine, that the perpetrator of the more gigantic fraud would require the stronger deterring motive to be kept straight. He is playing for heavy stakes, and we cannot hold out too strong a threat of infamy and suffering, if our aim is simply to prevent the crime. But neither, if we consider him from the purely moral point of view, would it be fair to argue that he was a better man than the pickpocket, because the plunder which tempted him was greater. The opposite, I fancy, would be true. He shows a callousness to human suffering, and an amount of deliberate hypocrisy and treachery which proves him to be not only the more dangerous, but the more thoroughly corrupt of the two. The two ends of providing a sufficient counterpoise and of punishing the worst men most severely, would, therefore, coincide in this case also; and the argument that the greater temptation implies less wickedness is plainly inapplicable.

Without going further into this, which may briefly indicate some of the perplexities involved, I may mention certain cases in which there seems to be a real divergence of the two principles. There are cases in which the temptation may be fairly held to lessen guilt, and in which punishment has, notwithstanding, been made severer in consequence. The criminal law of the last century, for example, imposed a penalty of death upon persons who stole certain kinds of property left in specially exposed positions. The ease of taking it would very possibly tempt to theft men who would elsewhere be honest; and it was sought to compensate for the strength of the temptation by more savage punishment of those who yielded to it. Or, again, there are certain problems of a similar kind connected with political offences. A man who gets up a rebellion from sincere political motives is generally far better morally than the man who gets up a rebellion for the sake, say, of simple plunder. Ought the motive to be allowed as an extenuation of the offence? It ought, it may be said, from a moral point of view; but, from the point of view of simple deterrence, we might rather consider that the patriotic rebel is the more dangerous person of the two, and, therefore, requires the prospect of at least as heavy a punishment to keep him quiet. So, again, it has been asked, whether it should be admitted as an excuse for a rioter, that he has joined in violent courses under threats from the riotous mob. This is, of course, an excuse from the moralist's point of view; the man is only attacking the police in order to save his own house from being burnt, not from a disorderly or disaffected spirit. But it is replied, from the deterring point of view, that, if such an excuse be allowed, you are ceasing to threaten at the precise moment when the threats are most required. If the law is not to press from one side, all the pressure will come from the other, and every argument will be in favour of joining the side of disorder. Hence, it is argued, we ought to proportion the punishment, not to the offence, but to the temptation.

Now, I may say, very briefly, that such a divergence of the two principles appears to me to be possible; and, further, that cases may be put in which it might be necessary to deter, at all hazards, even to the neglect of moral considerations. A general who is defending a town must sometimes burn the houses of innocent people, without stopping to consider whether they can ever be compensated; and I think that there may be analogous cases even in regard to law, where the consideration of the absolute necessity of putting down mischievous conduct may override the normal moral considerations. But the general answer is, I think, different, and may help to clear the principle. The law to which I have referred, for the protection of exposed property, obviously suggests one remark. The true remedy for the evil would have been not to increase the penalty, but to increase the protection. You ought to have provided more watchmen, or to have forbidden owners to put temptation in the way of their neighbours, and not to have tried to make the hangman do the work of the policeman. So our ancestors erred when they protected their fields, not by putting up fences, but by setting mantraps to mutilate occasional trespassers. In that, as in other cases, the mistake is to confuse between the deterring influence of punishment and the preventive influence of protective measures. Arguments, questionable when used on behalf of punishment considered as deterring, are perfectly applicable to the preventive measures. It is obviously right that such measures should be proportioned to the temptation. When a starving man steals a loaf, he is not so bad as a man who steals when he is not starving. We should, therefore, think it morally wrong to punish him as severely. But, if we thought that he ought not to have the loaf, we should take stronger precautions in proportion to the probable temptation. If, for example, we were sending supplies to relieve a starving district, it would be clearly right to send such a force with them as might prevent their appropriation by the strongest, or the first comers. But, at the same time, we should also think it right to save the men from temptation, by providing as much as possible against the danger of starvation. So, again, it would be monstrous to punish a poor man more severely than a duke, for stealing a watch; but, as a matter of prudence, I should take more precautions if I were dining in a poor public-house, than if I were dining in a ducal palace.

This suggests the true application of another doctrine, about the responsibility of society. Society, it is sometimes said, has no right to punish, because it ought to have suppressed the causes of crime. This doctrine is often stated very illogically, and would sanction a great deal of false sentimentalism. If society includes many corrupt and dangerous elements, that is no reason at all for not suppressing them by all available means. But, no doubt, it is a very good and sufficient reason for trying, as far as possible, to remove the cause as well as the effects; for getting rid of the temptations to crime, and training people so as to make them less disposed to crime, instead of simply punishing more severely those who have yielded to temptation and given play to instincts which have not been properly disciplined. This applies conspicuously to the case of the political criminal. It is generally essential to the welfare of a nation, that order should be preserved by a settled government. It is the duty of every government, not only to crush resistance, but to take such precautions as will make resistance hopeless. But a correlative duty is suggested when a rebellion actually occurs, and especially a rebellion which excites the sympathy of otherwise moral people. Such a case, that is, affords the strongest presumption that there are real grievances to be redressed, and that the rebel should not be confounded with the vulgar criminal. It may be, and often is, quite necessary to shoot him down, so long as he is actively attacking authority; but, when he is disarmed, he cannot be regarded simply as a thief or murderer, but as a man who has given a useful, though a disagreeable, hint that the times are out of joint.

I have gone so far into these questions—which might lead to a great many other problems of legal casuistry—with the desire of bringing out one essential part of the question. The difficulties which have arisen point, I think, to the impossibility of treating the problem exclusively, from a simple consideration of the deterring influence of punishment. That, however, remains an essential element. If the sole reason for punishing a sheep-stealer be not the prevention of sheep-stealing, that is, at least, a very excellent reason as far as it goes. But it seems to me an insufficient reason from the moral point of view, and, in particular, to fail in assigning a sufficiently distinct ground for determining the desirable degree of punishment. The principle was advocated as limiting the severity of the old laws; but it is not quite easy to define the limit suggested. There is a necessary clumsiness about the method. A punishment only becomes operative in the cases in which the threat has failed to deter. The fact that a man has committed a crime demonstrates the inadequacy of the system in his case; we have not given him a sufficient motive for abstaining. When Bentham says, that if a fine of five shillings would prevent a murder, you ought not to fine the murderer seven and sixpence, he says what is, in a sense, obviously true. If I could prevent a murder, or, indeed, achieve any other desirable object, for a given sum, why should I throw away another penny? But the fine is not inflicted till somebody has committed a murder, and, in that case, the threat of fining has obviously failed. The question arises, therefore, how far am I to go? Am I to go on raising the tariff till murder becomes altogether obsolete? But we have already got as far as capital punishment, without achieving that result. And, if we consider the case upon this method, we begin to find a difficulty in the method of calculation. We are to compare the pain inflicted upon the criminal with the pain saved to the victim. But the greater the pain inflicted, the smaller, according to the assumption made, will be the number of criminals, and the greater the number of victims saved. If we could adopt the Draconic system, and be sure of punishing every crime with death, crime ought to disappear; for hardly anybody would break the law if he were quite certain of the gallows. But, in that case, the pain, both of the criminal and the victim, would disappear, for there would be no one in either class. The result, therefore, would be a pure gain: no crime and no punishment. Against this practical conclusion, indeed, Bentham was one of the first to protest; and he uses one very sound argument. Punish all crime equally, he says, and you put a premium on the worst crimes. If both robber and murderer are to be hanged, the robber will have a good reason for destroying evidence, by adding the murder to the plunder of his victims. But, though the argument is very much to the purpose, it seems to make our calculations rather difficult. We cannot look simply to the deterring influence of a given punishment, but have to consider its place in the general tariff, and its influence in inducing people to prefer one variety of crime to another. And if we try to find our way out of this difficulty, we shall have, I think, to find that the mode of reasoning requires some modification.

The theory on which the calculation goes may, perhaps, be represented thus: It is supposed that by hanging a murderer, you prevent, say, ten murders which would otherwise happen. The suffering saved to the ten victims is greater than the actual suffering of the single criminal. Therefore, the infliction of the penalty gives a balance on the side of happiness. The argument seems to me to be sound as far as it goes, and, in some cases, it would, I think, be sufficient. If, for example, it were proved that the use of a certain remedy, such as inoculation, caused a certain number of deaths, while, on the other hand, it prevented ten times as many, we should consider that a good case had been made out for its adoption. And, similarly, if we attended simply to the number of executions and to the number of crimes, and could make the necessary arithmetical comparison, we should be able to estimate the balance of good or evil in terms of pain and pleasure. But this mode of considering the case is obviously inadequate; and, indeed, Bentham (though I cannot now go into his teaching) feels and makes allowance for its inadequacy. For, to say nothing else, the mere deterrence of a certain number of crimes is an entirely insufficient measure of the effect of the law. The one obvious remark is that, by suppressing violence, you not only save a certain number of lives, but you secure an essential condition of all civilised life. I came here to-night without a revolver in my pocket; and I am not aware that I showed any particular courage by doing so. But it would have been foolhardy to have shown the same negligence, a few years ago, in some of the Western States of America. If I had lived in such conditions, I should not only have taken a revolver, but have, very possibly, thought it a duty to join a vigilance committee, with a view to the suppression of crimes of violence. There are still regions where the fact that a man lives in a neighbouring village is a sufficient justification for shooting him down as soon as he comes in sight, for the simple reason that, otherwise, he would shoot you. So, when private war was still part of the regular custom, there was an obstacle which had to be crushed before any progress could be made in industrial development, which presupposes peaceful intercourse and mutual confidence. The formation of all that is meant by social order, the bringing about of a state of things in which men can meet habitually without fear or precaution, counting with complete confidence upon the absence of any hostile intention, is, obviously, an essential condition of everything that makes life worth living in a civilised country. The fact is too obvious to require much illustration; but it requires notice, for it is very imperfectly recognised when you regard murder, for example, simply as a kind of sporadic disease, which breaks out here and there, and can be kept within limits by killing some murderers, and so frightening other would-be murderers. The criminal law, no doubt, includes that consideration; but it includes infinitely more. It is a necessary corollary of that state of social relations which alone gives a secure base for every conceivable kind of satisfactory social relation. It might, perhaps, serve as a sufficient defence of the old system, when, in the absence of any settled order, the system of private vengeance, of blood-feuds, and so forth, served to restrain the prevalence of actual violence. But it is a totally insufficient measure of the real advantage gained by enforcing order. We have to compare, not only the number of murders and the number of victims which would exist in a given social order, supposing the penalty to be inflicted or not inflicted; but to compare two radically different social states, and to ask, whether it is better to live in a society where peace is the almost invariable rule, and violence the rare exception, or in one in which there is a chaos of little societies, each of them being in constant fear of all its neighbours. The construction of a central authority which will keep the peace is a necessary part of the process of civilisation, and the criminal law is involved in the process. For, of course, it follows that, so long as anti-social elements exist within the borders of society, and some people resort to the old methods of the knife or the bludgeon, they must be put down; and the hangman and the jailer, clumsy as the action may be, represent the only kind of machinery which has hitherto been invented for the purpose.

It follows that we must understand "deterrence" in a wider sense than we have hitherto given to it. When we speak of punishment as deterring from crime, we must consider, not merely the effect upon the individual of the prospect of punishment following detection, but the total effect of a systematic adherence to the law upon the preservation of a peaceful state of society at large. We do not simply wish to provide a sufficient motive to decide the individual who is asking himself, shall I steal or not steal? but to maintain an organisation under which property shall be normally respected, and stealing become as exceptional as we can make it. This, in turn, involves much more than a simple execution of the criminal law; it involves the support of agencies for prevention, education, and reformation; though it does, also, involve an inflexible adherence to the criminal law. The law has to use rough means, and cannot possibly affect to adhere precisely to the moral deserts of individual cases. But it is justified by the simple ground that the only alternative is a chaos of barbarism. If you ask, therefore, in what sense is a criminal law just? we must confess that, in certain respects, it is impossible that it should be strictly just; it must deal with the found-out exclusively and with those who are found out in certain definite cases of criminality, and it must, therefore, impose penalties which do not precisely correspond to the degree of criminality implied. But the relation to morality is, nevertheless, intimate. For the growth of the social order depends upon the growth of the corresponding social instincts; or rather, the two processes are correlative. If I love my neighbour I shall not wish to cut his throat; and, in order that I may love him, I must be pretty sure that he does not mean to cut mine. The external framework provides a protection under which the primary moral instincts can expand; and the expansion of the instincts supposes a correlative modification of the external framework. The moral requirement in regard to the criminal law is, therefore, essentially, that it should be such a law as is favourable, when considered in connection with the whole order, to the strength and development of the existing morality. If the criminal asks, How do you justify yourself for punishing me? the reply must be, Because the inflexible administration of the law is an essential precondition of the whole system, under which alone progress is possible. A society in which peace and order are preserved is superior, in morals as in other respects, to a society in which peace and order are made impossible by violence; and the suppression by punishment of offenders is involved in the system. The advantage of belonging to such a society is not to be measured by counting up the working of individual cases; but by the whole characteristics of the social state, taken as a whole, and including, as one essential part, the administration of criminal law in such a way as to be in conformity with the conditions of healthy social development. The difficulty, I think, though I can only indicate the argument briefly, results from a common illusion, which is illustrated by the once famous social contract theory. You suppose a number of independent individuals, agreeing to join and expecting to receive a precise equivalent for every sacrifice that they make in consequence. The reply is, that the individual is the product of the society, and it is a mere fiction to consider him as possessing any antecedent rights whatever. His rights are to be deduced from, not to supply the premisses for deducing, the social order. The only considerations which are relevant are those which affect the welfare of the social organism, taken as a whole; and we must regard them as determined, before we come to the distribution of benefits and burdens among its constituent facts. Otherwise, we should be falling into the same fallacy as if we argued about the health of separate bodily organs, legs, and arms, and stomachs, as though they were independent things, fastened together to make a single machine. Since the leg implies the stomach, any consideration of the leg's separate rights would be absurd. So the individual member of a political society cannot be regarded as though he had existed outside society somewhere, and was entitled to a precise equivalent for the sacrifice of his independence. The doctrine involves impossible considerations. I have to contribute to certain sanitary regulations, though I may be stronger or weaker than my neighbours, and therefore less or more in need of them. Or, I have to pay a school-rate, whether I have a dozen children or none at all. Do those facts give me a right to complain if I am taxed equally with my neighbours? If so, every benefit which I receive from society must be set down as a separate item in an account to be balanced by itself. Obviously, the advantage which I receive in such cases is the whole advantage received from living in a healthy place or among educated people; and it is essentially impossible to cut that up into a number of different bits of happiness conferred in return for separate payments on account. If I use the contract formula, I must interpret it to mean that amenability to various regulations, including the criminal law, is part of the whole bargain, which would have been made, if it had ever been real, when I decided, if I ever had decided, to join the society. The instinct for punishing criminals guilty of violence is one of the fundamental instincts of civilisation, and we must accept it just as we accept any other fundamental instinct.

The question of justice, however, is not a whit the less essential because it presupposes this social characteristic instead of supplying the primary axioms from which it is to be deduced. It is undoubtedly of the highest importance that every difference in our method of treating different classes should have its sufficient reason, to be assigned as clearly as possible. The preservation of the peace is essential; but that does not settle the methods by which it is to be preserved.

On what ground, then, are we to deal with the problem of justice as regards different classes of crime? If the calculation of pain and pleasure, as already stated, seems to be unsatisfactory, what is the right principle of proportioning punishment to offence? I have noticed one argument which Bentham applied, and, as I think, with very good reason. To punish crimes equally, he said, is virtually to put a premium upon the worst. The "in for a penny in for a pound" maxim becomes at once applicable. Moreover, as every one now admits, the old brutal system is condemned by experience. To punish a great number of offences with death led to a mixture of excessive brutality with excessive uncertainty. The cruel punishment of some criminals was balanced by the complete escape of others. But this practical failure clearly resulted, in great measure, from an obscure sense of justice. It was grossly unjust, it seemed, to hang a man for stealing a loaf, when you could only hang another for the brutal murder of his wife. The penalty in the first case, was, it was felt, altogether out of proportion to the offence. This instinctive sentiment was, as I think we all feel, substantially right. In any case, it would have to be taken into account by the legislator, for the obvious reason that punishments which outrun public opinion, tend to make martyrs of criminals. They are either not inflicted, or they set the sympathy of the people on the side of the offender. But to say this, is not to prove the sentiment to be just, only to take account of its existence. And the question, therefore, remains, how it is to be logically justified, for it may seem to imply the theory to which I have objected—the hypothesis of a sort of debtor and creditor account—of the old "eye for an eye" doctrine, which, as I have argued, involves a misconception of the true doctrine. My reply would be, in general terms, that the doctrine requires restatement, and, if properly stated, will not lose but acquire new forces.

Let us consider the consequences of my previous statements. The essential condition of social development is enforcement, where necessary, of peace and order by adequate means. The criminal law corresponds to one part of this process. The whole social system includes machinery for prevention, for reformation and for education, as well as for punishment; and it is only when taken in its relation to other parts of the system, that we can give the full justification. Its methods are, as I have said, obviously full of imperfections, from the purely moral point of view. If we consider it as an isolated fact, comparably to the interference of a quasi-supernatural power, which clutches an offender here and there, and punishes him simply to frighten others, the arbitrary and unequal nature of the proceeding assumes an air of injustice. In fact, if you take the extreme individualist view, according to which each man is an independent unit, while society represents a force impinging upon him from without, it always becomes difficult to introduce the conception of justice without ending in the approval of anarchy. When, however, we consider the social organisation as including all the means of civilising society, of strengthening the general spirit of order, as well as acting upon the fears of the disorderly, we have to take wider considerations into account. We become sensible, in the first place, of the importance of the principle that punishment should never be substituted for prevention. Wherever it is possible to remove temptations, or take precautions which make crime impossible, we can have no excuse for adopting the blundering and unsatisfactory system of punishing those who have committed it. We admit, that is, that the criminal law, though absolutely necessary, is an essentially clumsy contrivance, to be used only when other methods fail. When certain punishments have been condemned as brutalising, it has been replied that the persons punished were already so brutal that it is impossible to make them worse. But the brutalising influence is even more objectionable as it applies to the legislator than as it applies to the criminal. To make up for neglect of appropriate precautions by severity against the offender, is to adopt the necessarily arbitrary method in which chance must always play a part in place of more effective and civilising methods. Frugality in applying punishment is desirable as a guarantee that we are acting in the proper spirit. An Indian official was asked why the native police were disposed to use torture for the detection of crime. The cause was, he said, mainly from laziness: it was so much easier to sit in the shade, rubbing red pepper in a poor devil's eyes, than to go about in a hot sun collecting evidence. So, it would be very much easier to inflict cruel punishment than to try to remove the causes of crime; and a resolution never to use the more brutal methods is not, as I think, to be regarded as a proof of weak sentimentalism, but as a judicious self-denying ordinance, imposed upon society by itself, as binding it always to adopt, as far as it possibly can, what is at once the more humane and the more scientific method. The same principle involves the careful graduation of punishment. There are, indeed, as I believe, though I cannot give reasons, cases in which crimes ought to be punished with death. There are persons of whom we may say that it would have been better, especially for their neighbours, if they had never been born. "I am worth inconceivably more for hanging than for any other purpose," said the heroic John Brown; and the words may be applied, in a very different sense, to some of the wretches who occasionally make their appearance in the courts. To hang such a man is to act upon the assumption that murderers represent elements which are entirely and radically anti-social. The only remedy for them is extirpation. But, if this be admitted, it suggests a sufficient reason for not applying it to the cases of less gravity, in which such radical incompatibility has not been demonstrated. Punishment by death, even if necessary, is certainly a confession of impotence. We are admitting that we can do nothing better with the man than convert him into a scarecrow for the benefit of his like. What more, it may be asked, can we do with a criminal? The obvious reply would be, reform him. Although no one can doubt that reformation would be an extremely good thing, wherever practicable, it may be urged that the enterprise is exceedingly difficult; that, in many cases, it is hopeless; and that we might spend our money and our efforts to better purpose upon more hopeful materials. And yet, I think that the answer is the true one, if properly understood, and will suggest the right meaning to be given to the word "deterrence". So long as we consider the individual case alone, and merely mean that we are giving motives to bad men for refraining from particular lines of conduct, the results, however desirable, are of limited value. But if we consider deterrence as including or coinciding with reformation, as indicating a part of the general system of moral pressure by which the classes exposed to temptation may be gradually raised in the scale of civilisation, we recognise an acceptable meaning. In fact, if we ask what is the deterring influence of punishment, we must observe that at one extreme it will always fail, or only induce a bad man to take precautions against detection; and that, at the other end of the scale, there are a great many cases in which it does not come into active operation at all. You and I, I hope, are not in the least disposed to assault each other, even though no policeman is present. The bare thought of resorting to violence, pelting me, say, with rotten eggs, has not even suggested itself to you, even though I may be making a very provoking use of my tongue. But there is also an intermediate class of people upon whom the possibility of having to appear in a police court, and the strong sense of shame attached to such appearances, is an active restraining force, tending to limit, and, in cases where the proper conditions exist, gradually to narrow, the sphere of violence. We, the peaceable and law-abiding citizens, have gained a right to those epithets, because we have lived in a sphere where the law has been habitually enforced. We have ceased to carry deadly weapons about us, and have established a general condition of good order. The deterring influence of the criminal law acts, or ought to act, by gradually spreading that state of mind through a steadily widening circle. The classes which are still in need of such a support to their moral instincts are clearly capable of reformation, whatever may be the case of some of the individuals who break the law. A fighting tribe, which has been in the habit of resenting every injury by the use of the knife, may learn, in a very short time, that a court of law settles disputes more agreeably than a free fight; and may become a most admirable and efficient part of the society to which it belongs. And the same may be said of large classes in our own society, which are perfectly capable of being converted into good citizens, though they may retain certain propensities developed under a rougher and more brutal system. To employ excessive and brutalising punishments in order to suppress small offences, is, therefore, to abandon the aim of civilising, to declare internecine war against the class, and to regard them simply as a nuisance to be abated. The effect might be, if the law could be carried out, to prevent a certain number of crimes; but it must also be to generate a more dangerous spirit in the class which you regard simply as dangerous, instead of regarding it as the possible raw materials of a more civilised and orderly society. Without attempting to dwell upon a familiar argument, I merely say that this view of the case implies that the governing power should be regarded, not simply as a machinery for catching and killing noxious criminals, but as a great civilising influence, suppressing all temptations to crime, where possible; preferring prevention, in every practicable case, to punishment, and making use of the clumsy, though necessary, weapons in the last resort; and acting by a steady and regulated pressure upon all anti-social elements. It is only possible to give a satisfactory theory of the jail and the gallows, when you take them as a subordinate part of the system which includes reformatories and schools, and due precautions for the regular preservation of order. The ultimate criterion of justice is not to be found in any attempt to form a debtor and creditor account between the government and the individual; but in the civilising influence of the system, taken as a whole.

And, finally, I come back to the other theory which I have noticed. To supply the defects of the simply deterrent theory, it has been found necessary, as I said, to invoke the vindictive theory. We should go, it was suggested, upon the theory that a criminal is hateful, and, therefore, that it should be a pleasure to punish him. The feelings of resentment and moral indignation are parts of our nature, to which the punishment of the offender affords them a legitimate gratification. Now, to this, I should reply that, in the first place, I do not admit that the desire for revenge, as usually understood, can ever be legitimate. Revenge, as I understand the word, implies a personal feeling. It is taking pleasure in giving pain to a man because he has given pain to me. According to my view of morals, any pleasure in causing pain is, so far, wrong; and the public punishment should be free from all personal motive. I quite agree with Bentham that we ought not to take a positive pleasure in the sufferings, even of the worst criminal; and to admit the legitimacy of such pleasure is to admit an element of pure sentiment to which it is difficult to assign any precise limits. If you allow yourself to hate a man so as to take pleasure in his sufferings, you might justify the infliction of superfluous torture and the old methods of hanging, drawing, and quartering. To do so is precisely to approve the ferocious old treatment, to which, as I conceive, the theory of simple deterrence was an excellent corrective, in so far as it at least implied a definite limit to the indulgence of fiercer passions. There is, however, I think, an element of truth in the doctrine. I admit, that is, that the punishment of a criminal should carry a moral approval, and not be regarded purely as a measure of convenience. Successful crime should be regarded with abhorrence. If a man convicted of a grave offence should be allowed to go without punishment, we should be rightly aggrieved. It is not, however, that we should take pleasure in his suffering, but that we should be pained by an example of the practical impunity of anti-social conduct. The escape of a murderer would, as we should feel, be a blow to the security of all innocent people. In that sense, we may take pleasure in his punishment, not in the sense of positive enjoyment, but, certainly, in the sense of relief from positive sense of evil. It is, and should be, painful to see the rogues flourish and honest men droop, and to observe "captive good attending captain ill". But the pleasure of seeing the necessary equilibrium restored is different from the pleasure of dwelling upon the sufferings of the disturber. The practical difference is that, while we regard the infliction of suffering as necessary, we admit it to be a necessary evil, and are keenly alive to the inability of keeping it within the limits fixed by the general necessities of the law.